v21.6.0

.gitignore

@@ -11,5 +11,6 @@ gui-user.bat
 gui-user.ps1
 .vscode
 wandb
+setup.log
 logs
-SmilingWolf
+SmilingWolf

README.md

@@ -345,6 +345,17 @@ This will store a backup file with your current locally installed pip packages a
 
 ## Change History
 
+* 2023/06/02 (v21.6.0)
+- Merge kohya_ss repo changes
+- Improge logging of kohya_ss GUI
+* 2023/05/28 (v21.5.15)
+- Show warning when image caption file does not exist during training. [PR #533](https://github.com/kohya-ss/sd-scripts/pull/533) Thanks to TingTingin!
+  - Warning is also displayed when using class+identifier dataset. Please ignore if it is intended.
+- `train_network.py` now supports merging network weights before training. [PR #542](https://github.com/kohya-ss/sd-scripts/pull/542) Thanks to u-haru!
+  - `--base_weights` option specifies LoRA or other model files (multiple files are allowed) to merge.
+  - `--base_weights_multiplier` option specifies multiplier of the weights to merge (multiple values are allowed). If omitted or less than `base_weights`, 1.0 is used.
+  - This is useful for incremental learning. See PR for details.
+- Show warning and continue training when uploading to HuggingFace fails.
 * 2023/05/28 (v21.5.14)
 - Add Create Groupo tool and GUI
 * 2023/05/24 (v21.5.13)
@@ -391,69 +402,3 @@ This will store a backup file with your current locally installed pip packages a
     - The actual value of the noise offset is calculated as `noise_offset + abs(mean(latents, dim=(2,3))) * adaptive_noise_scale`. Since the latent is close to a normal distribution, it may be a good idea to specify a value of about 1/10 to the same as the noise offset.
     - Negative values can also be specified, in which case the noise offset will be clipped to 0 or more.
   - Other minor fixes.
-* 2023/04/06 (v21.5.9)
-  - Inplement headless mode to enable easier support under headless services like vast.ai. To make use of it start the gui with the `--headless` argument like:
-
-    `.\gui.ps1 --headless` or `.\gui.bat --headless` or `./gui.sh --headless`
-  - Added the option for the user to put the wandb api key in a textbox under the advanced configuration dropdown and a checkbox to toggle for using wandb logging. @x-CK-x
-  - Docker build image @Trojaner
-    - Updated README to use docker compose run instead of docker compose up to fix broken tqdm
-      - Related: Doesn't work with docker-compose tqdm/tqdm#771
-    - Fixed build for latest release
-    - Replace pillow with pillow-simd
-    - Removed --no-cache again as pip cache is not enabled anyway
-  - While overwriting .txt files with prefix and postfix including different encodings you might encounter this decoder error. This small fix gets rid of it... @ertugrul-dmr
-  - Docker Add --no-cache-dir to reduce image size @chiragjn
-  - Reverting bitsandbytes version to 0.35.0 due to issues with 0.38.1 on some systems
-* 2023/04/05 (v21.5.8)
-  - Add `Cache latents to disk` option to the gui.
-  - When saving v2 models in Diffusers format in training scripts and conversion scripts, it was found that the U-Net configuration is different from those of Hugging Face's stabilityai models (this repository is `"use_linear_projection": false`, stabilityai is `true`). Please note that the weight shapes are different, so please be careful when using the weight files directly. We apologize for the inconvenience.
-      - Since the U-Net model is created based on the configuration, it should not cause any problems in training or inference.
-      - Added `--unet_use_linear_projection` option to `convert_diffusers20_original_sd.py` script. If you specify this option, you can save a Diffusers format model with the same configuration as stabilityai's model from an SD format model (a single `*.safetensors` or `*.ckpt` file). Unfortunately, it is not possible to convert a Diffusers format model to the same format.
-  - Lion8bit optimizer is supported. [PR #447](https://github.com/kohya-ss/sd-scripts/pull/447) Thanks to sdbds!
-    - Currently it is optional because you need to update `bitsandbytes` version. See "Optional: Use Lion8bit" in installation instructions to use it.
-  - Multi-GPU training with DDP is supported in each training script. [PR #448](https://github.com/kohya-ss/sd-scripts/pull/448) Thanks to Isotr0py!
-  - Multi resolution noise (pyramid noise) is supported in each training script. [PR #471](https://github.com/kohya-ss/sd-scripts/pull/471) Thanks to pamparamm!
-    - See PR and this page [Multi-Resolution Noise for Diffusion Model Training](https://wandb.ai/johnowhitaker/multires_noise/reports/Multi-Resolution-Noise-for-Diffusion-Model-Training--VmlldzozNjYyOTU2) for details.
-  - Add --no-cache-dir to reduce image size @chiragjn
-* 2023/05/01 (v21.5.7)
-  - `tag_images_by_wd14_tagger.py` can now get arguments from outside. [PR #453](https://github.com/kohya-ss/sd-scripts/pull/453) Thanks to mio2333!
-  - Added `--save_every_n_steps` option to each training script. The model is saved every specified steps.
-    - `--save_last_n_steps` option can be used to save only the specified number of models (old models will be deleted).
-    - If you specify the `--save_state` option, the state will also be saved at the same time. You can specify the number of steps to keep the state with the `--save_last_n_steps_state` option (the same value as `--save_last_n_steps` is used if omitted).
-    - You can use the epoch-based model saving and state saving options together.
-    - Not tested in multi-GPU environment. Please report any bugs.
-  - `--cache_latents_to_disk` option automatically enables `--cache_latents` option when specified. [#438](https://github.com/kohya-ss/sd-scripts/issues/438)
-  - Fixed a bug in `gen_img_diffusers.py` where latents upscaler would fail with a batch size of 2 or more.
-  - Fix triton error
-  - Fix issue with merge lora path with spaces
-  - Added support for logging to wandb. Please refer to PR #428. Thank you p1atdev!
-    - wandb installation is required. Please install it with pip install wandb. Login to wandb with wandb login command, or set --wandb_api_key option for automatic login.
-    - Please let me know if you find any bugs as the test is not complete.
-  - You can automatically login to wandb by setting the --wandb_api_key option. Please be careful with the handling of API Key. PR #435 Thank you Linaqruf!
-  - Improved the behavior of --debug_dataset on non-Windows environments. PR #429 Thank you tsukimiya!
-  - Fixed --face_crop_aug option not working in Fine tuning method.
-  - Prepared code to use any upscaler in gen_img_diffusers.py.
-  - Fixed to log to TensorBoard when --logging_dir is specified and --log_with is not specified.
-  - Add new docker image solution.. Thanks to @Trojaner 
-* 2023/04/22 (v21.5.5)
-    - Update LoRA merge GUI to support SD checkpoint merge and up to 4 LoRA merging
-    - Fixed `lora_interrogator.py` not working. Please refer to [PR #392](https://github.com/kohya-ss/sd-scripts/pull/392) for details. Thank you A2va and heyalexchoi!
-    - Fixed the handling of tags containing `_` in `tag_images_by_wd14_tagger.py`.
-    - Add new Extract DyLoRA gui to the Utilities tab.
-    - Add new Merge LyCORIS models into checkpoint gui to the Utilities tab.
-    - Add new info on startup to help debug things
-* 2023/04/17 (v21.5.4)
-    - Fixed a bug that caused an error when loading DyLoRA with the `--network_weight` option in `train_network.py`.
-    - Added the `--recursive` option to each script in the `finetune` folder to process folders recursively. Please refer to [PR #400](https://github.com/kohya-ss/sd-scripts/pull/400/) for details. Thanks to Linaqruf!
-    - Upgrade Gradio to latest release
-    - Fix issue when Adafactor is used as optimizer and LR Warmup is not 0: https://github.com/bmaltais/kohya_ss/issues/617
-    - Added support for DyLoRA in `train_network.py`. Please refer to [here](./train_network_README-ja.md#dylora) for details (currently only in Japanese).
-    - Added support for caching latents to disk in each training script. Please specify __both__ `--cache_latents` and `--cache_latents_to_disk` options.
-        - The files are saved in the same folder as the images with the extension `.npz`. If you specify the `--flip_aug` option, the files with `_flip.npz` will also be saved.
-        - Multi-GPU training has not been tested.
-        - This feature is not tested with all combinations of datasets and training scripts, so there may be bugs.
-    - Added workaround for an error that occurs when training with `fp16` or `bf16` in `fine_tune.py`.
-    - Implemented DyLoRA GUI support. There will now be a new 'DyLoRA Unit` slider when the LoRA type is selected as `kohya DyLoRA` to specify the desired Unit value for DyLoRA training.
-    - Update gui.bat and gui.ps1 based on: https://github.com/bmaltais/kohya_ss/issues/188
-    - Update `setup.bat` to install torch 2.0.0 instead of 1.2.1. If you want to upgrade from 1.2.1 to 2.0.0 run setup.bat again, select 1 to uninstall the previous torch modules, then select 2 for torch 2.0.0

dreambooth_gui.py

@@ -40,6 +40,11 @@
 from library.utilities import utilities_tab
 from library.sampler_gui import sample_gradio_config, run_cmd_sample
 
+from library.custom_logging import setup_logging
+
+# Set up logging
+log = setup_logging()
+
 # from easygui import msgbox
 
 folder_symbol = '\U0001f4c2'  # 📂
@@ -129,14 +134,14 @@ def save_configuration(
     save_as_bool = True if save_as.get('label') == 'True' else False
 
     if save_as_bool:
-        print('Save as...')
+        log.info('Save as...')
         file_path = get_saveasfile_path(file_path)
     else:
-        print('Save...')
+        log.info('Save...')
         if file_path == None or file_path == '':
             file_path = get_saveasfile_path(file_path)
 
-    # print(file_path)
+    # log.info(file_path)
 
     if file_path == None or file_path == '':
         return original_file_path  # In case a file_path was provided and the user decide to cancel the open action
@@ -253,7 +258,7 @@ def open_configuration(
         # load variables from JSON file
         with open(file_path, 'r') as f:
             my_data = json.load(f)
-            print('Loading config...')
+            log.info('Loading config...')
             # Update values to fix deprecated use_8bit_adam checkbox and set appropriate optimizer if it is set to True
             my_data = update_my_data(my_data)
     else:
@@ -407,7 +412,7 @@ def train_model(
 
     # Check if subfolders are present. If not let the user know and return
     if not subfolders:
-        print(
+        log.info(
             '\033[33mNo subfolders were found in',
             train_data_dir,
             " can't train\...033[0m",
@@ -422,7 +427,7 @@ def train_model(
         try:
             repeats = int(folder.split('_')[0])
         except ValueError:
-            print(
+            log.info(
                 '\033[33mSubfolder',
                 folder,
                 "does not have a proper repeat value, please correct the name or remove it... can't train...\033[0m",
@@ -444,30 +449,30 @@ def train_model(
         )
 
         if num_images == 0:
-            print(f'{folder} folder contain no images, skipping...')
+            log.info(f'{folder} folder contain no images, skipping...')
         else:
             # Calculate the total number of steps for this folder
             steps = repeats * num_images
             total_steps += steps
 
             # Print the result
-            print('\033[33mFolder', folder, ':', steps, 'steps\033[0m')
+            log.info('\033[33mFolder', folder, ':', steps, 'steps\033[0m')
 
     if total_steps == 0:
-        print(
+        log.info(
             '\033[33mNo images were found in folder',
             train_data_dir,
             '... please rectify!\033[0m',
         )
         return
 
     # Print the result
-    # print(f"{total_steps} total steps")
+    # log.info(f"{total_steps} total steps")
 
     if reg_data_dir == '':
         reg_factor = 1
     else:
-        print(
+        log.info(
             '\033[94mRegularisation images are used... Will double the number of steps required...\033[0m'
         )
         reg_factor = 2
@@ -482,7 +487,7 @@ def train_model(
             * int(reg_factor)
         )
     )
-    print(f'max_train_steps = {max_train_steps}')
+    log.info(f'max_train_steps = {max_train_steps}')
 
     # calculate stop encoder training
     if int(stop_text_encoder_training_pct) == -1:
@@ -493,10 +498,10 @@ def train_model(
         stop_text_encoder_training = math.ceil(
             float(max_train_steps) / 100 * int(stop_text_encoder_training_pct)
         )
-    print(f'stop_text_encoder_training = {stop_text_encoder_training}')
+    log.info(f'stop_text_encoder_training = {stop_text_encoder_training}')
 
     lr_warmup_steps = round(float(int(lr_warmup) * int(max_train_steps) / 100))
-    print(f'lr_warmup_steps = {lr_warmup_steps}')
+    log.info(f'lr_warmup_steps = {lr_warmup_steps}')
 
     run_cmd = f'accelerate launch --num_cpu_threads_per_process={num_cpu_threads_per_process} "train_db.py"'
     if v2:
@@ -606,7 +611,7 @@ def train_model(
         output_dir,
     )
 
-    print(run_cmd)
+    log.info(run_cmd)
 
     # Run the command
     if os.name == 'posix':
@@ -979,11 +984,11 @@ def UI(**kwargs):
     css = ''
 
     headless = kwargs.get('headless', False)
-    print(f'headless: {headless}')
+    log.info(f'headless: {headless}')
 
     if os.path.exists('./style.css'):
         with open(os.path.join('./style.css'), 'r', encoding='utf8') as file:
-            print('Load CSS...')
+            log.info('Load CSS...')
             css += file.read() + '\n'
 
     interface = gr.Blocks(

finetune_gui.py

@@ -31,6 +31,11 @@
 from library.utilities import utilities_tab
 from library.sampler_gui import sample_gradio_config, run_cmd_sample
 
+from library.custom_logging import setup_logging
+
+# Set up logging
+log = setup_logging()
+
 # from easygui import msgbox
 
 folder_symbol = '\U0001f4c2'  # 📂
@@ -128,14 +133,14 @@ def save_configuration(
     save_as_bool = True if save_as.get('label') == 'True' else False
 
     if save_as_bool:
-        print('Save as...')
+        log.info('Save as...')
         file_path = get_saveasfile_path(file_path)
     else:
-        print('Save...')
+        log.info('Save...')
         if file_path == None or file_path == '':
             file_path = get_saveasfile_path(file_path)
 
-    # print(file_path)
+    # log.info(file_path)
 
     if file_path == None or file_path == '':
         return original_file_path  # In case a file_path was provided and the user decide to cancel the open action
@@ -258,7 +263,7 @@ def open_configuration(
         # load variables from JSON file
         with open(file_path, 'r') as f:
             my_data = json.load(f)
-            print('Loading config...')
+            log.info('Loading config...')
             # Update values to fix deprecated use_8bit_adam checkbox and set appropriate optimizer if it is set to True
             my_data = update_my_data(my_data)
     else:
@@ -391,7 +396,7 @@ def train_model(
         if full_path:
             run_cmd += f' --full_path'
 
-        print(run_cmd)
+        log.info(run_cmd)
 
         # Run the command
         if os.name == 'posix':
@@ -416,7 +421,7 @@ def train_model(
         if full_path:
             run_cmd += f' --full_path'
 
-        print(run_cmd)
+        log.info(run_cmd)
 
         # Run the command
         if os.name == 'posix':
@@ -433,10 +438,10 @@ def train_model(
             if lower_f.endswith(('.jpg', '.jpeg', '.png', '.webp'))
         ]
     )
-    print(f'image_num = {image_num}')
+    log.info(f'image_num = {image_num}')
 
     repeats = int(image_num) * int(dataset_repeats)
-    print(f'repeats = {str(repeats)}')
+    log.info(f'repeats = {str(repeats)}')
 
     # calculate max_train_steps
     max_train_steps = int(
@@ -452,10 +457,10 @@ def train_model(
     if flip_aug:
         max_train_steps = int(math.ceil(float(max_train_steps) / 2))
 
-    print(f'max_train_steps = {max_train_steps}')
+    log.info(f'max_train_steps = {max_train_steps}')
 
     lr_warmup_steps = round(float(int(lr_warmup) * int(max_train_steps) / 100))
-    print(f'lr_warmup_steps = {lr_warmup_steps}')
+    log.info(f'lr_warmup_steps = {lr_warmup_steps}')
 
     run_cmd = f'accelerate launch --num_cpu_threads_per_process={num_cpu_threads_per_process} "./fine_tune.py"'
     if v2:
@@ -560,7 +565,7 @@ def train_model(
         output_dir,
     )
 
-    print(run_cmd)
+    log.info(run_cmd)
 
     # Run the command
     if os.name == 'posix':
@@ -931,11 +936,11 @@ def UI(**kwargs):
     css = ''
 
     headless = kwargs.get('headless', False)
-    print(f'headless: {headless}')
+    log.info(f'headless: {headless}')
 
     if os.path.exists('./style.css'):
         with open(os.path.join('./style.css'), 'r', encoding='utf8') as file:
-            print('Load CSS...')
+            log.info('Load CSS...')
             css += file.read() + '\n'
 
     interface = gr.Blocks(

gui.bat

@@ -5,7 +5,7 @@ call .\venv\Scripts\activate.bat
 set PATH=%PATH%;%~dp0venv\Lib\site-packages\torch\lib
 
 :: Debug info about system
-python.exe .\tools\debug_info.py
+:: python.exe .\tools\debug_info.py
 
 :: Validate the requirements and store the exit code
 python.exe .\tools\validate_requirements.py

gui.ps1

@@ -3,7 +3,7 @@
 $env:PATH += ";$($MyInvocation.MyCommand.Path)\venv\Lib\site-packages\torch\lib"
 
 # Debug info about system
-python.exe .\tools\debug_info.py
+# python.exe .\tools\debug_info.py
 
 # Validate the requirements and store the exit code
 python.exe .\tools\validate_requirements.py
@@ -16,6 +16,6 @@ if ($LASTEXITCODE -eq 0) {
         $argsFromFile = Get-Content .\gui_parameters.txt -Encoding UTF8 | Where-Object { $_ -notmatch "^#" } | Foreach-Object { $_ -split " " }
     }
     $args_combo = $argsFromFile + $args
-    Write-Host "The arguments passed to this script were: $args_combo"
+    # Write-Host "The arguments passed to this script were: $args_combo"
     python.exe kohya_gui.py $args_combo
 }